Multi-Modal Face Authentication with IR Liveness Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing authentication methods, such as face recognition, can lead to unwanted auto-login in the presence of others and can be compromised by unauthorized use of user photos, lacking multi-modal verification to ensure secure access.

Innovation Solution

A computing device employs a processor to execute behaviors only after receiving a combination of inputs, including face recognition, infrared sensing, user-defined facial expressions, or physical gestures, allowing users to select and define multi-modal authentication protocols ensuring secure access.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If face recognition is used for authentication, then authentication speed is improved, but security deteriorates due to unwanted auto-login and photo compromise

Engineering Contradiction:
Improveauthentication speedVSAvoidauthentication security
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The patent combines face recognition with infrared sensing and facial expression detection into a multi-modal authentication system. The processor requires multiple input types (visible face image, IR signal magnitude, facial expression match) to be satisfied simultaneously, creating a unified authentication mechanism that maintains speed while improving security through layered verification

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces infrared sensing as an intermediary layer between the visible face recognition and the final authentication decision. The IR sensor detects thermal radiation patterns that serve as a mediator to verify liveness, preventing photo-based attacks while maintaining the convenience of face-based authentication

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If multiple input modes are required for authentication, then security is improved, but device complexity increases

Engineering Contradiction:
Improveauthentication securityVSAvoidauthentication system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent makes the existing camera and sensor components perform multiple functions: the camera captures both visible light images for face recognition and infrared signals for thermal detection, while also detecting facial expressions. This multi-functionality approach increases security through multiple verification layers without requiring separate dedicated hardware for each authentication modality

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent merges the processing of multiple authentication inputs (face image matching, IR signal threshold checking, facial expression recognition) into a single integrated processor decision logic. The processor evaluates all inputs simultaneously and makes one unified authentication determination, reducing operational complexity despite the multi-modal nature of the system

Inventive Principle:
Principle #5Merging (Combining)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enhances security by requiring multiple verified inputs, preventing unauthorized access and ensuring that only the authorized user can unlock the device, even when images are used for authentication.

Implementation Method 1

face recognition in combination with infrared (IR) sensing

Methodology Applied
Scientific EffectInfrared sensing: Infrared Radiation

Data Source

PatentEP2709031B1Gesture- and expression-based authentication
Publication Date: 2020.03.25 SONY GROUP CORP
  • EP2709031B1 patent drawingFigure 1
  • EP2709031B1 patent drawingFigure 2
  • EP2709031B1 patent drawingFigure 3~4

AI summary

A user can define a gesture-based input mode with respective input value to establish an authentication protocol to unlock a computer or govern other computer behavior. As an alternative or in addition, the user can define a second input mode based on face recognition plus IR sensing satisfying a threshold to ensure a live person is being imaged for authentication, and/or face recognition plus a particular facial expression such as a smile and wink.